DataTransferItem: webkitGetAsEntry() method

None.

A FileSystemEntry-based object describing the dropped item. This will be either FileSystemFileEntry or FileSystemDirectoryEntry. The method aborts and returns null if the dropped item isn't a file, or if the DataTransferItem object is not in read or read/write mode.

In this example, a drop zone is created, which responds to the drop event by scanning through the dropped files and directories, outputting a hierarchical directory listing.

The HTML establishes the drop zone itself, which is a <div> element with the ID "dropzone", and an unordered list element with the ID "listing".

<p>Drag files and/or directories to the box below!</p>

<div id="dropzone">
  <div id="boxtitle">Drop Files Here</div>
</div>

<h2>Directory tree:</h2>

<ul id="listing"></ul>

The styles used by the example are shown here.

#dropzone {
  text-align: center;
  width: 300px;
  height: 100px;
  margin: 10px;
  padding: 10px;
  border: 4px dashed red;
  border-radius: 10px;
}

#boxtitle {
  display: table-cell;
  vertical-align: middle;
  text-align: center;
  color: black;
  font: bold 2em "Arial", sans-serif;
  width: 300px;
  height: 100px;
}

body {
  font: 14px "Arial", sans-serif;
}

First, let's look at the recursive scanFiles() function. This function takes as input a FileSystemEntry representing an entry in the file system to be scanned and processed (the item parameter), and an element into which to insert the list of contents (the container parameter).

let dropzone = document.getElementById("dropzone");
let listing = document.getElementById("listing");

function scanFiles(item, container) {
  let elem = document.createElement("li");
  elem.textContent = item.name;
  container.appendChild(elem);

  if (item.isDirectory) {
    let directoryReader = item.createReader();
    let directoryContainer = document.createElement("ul");
    container.appendChild(directoryContainer);
    directoryReader.readEntries((entries) => {
      entries.forEach((entry) => {
        scanFiles(entry, directoryContainer);
      });
    });
  }
}

scanFiles() begins by creating a new <li> element to represent the item being scanned, inserts the name of the item into it as its text content, and then appends it to the container. The container is always a list element in this example, as you'll see shortly.

Once the current item is in the list, the item's isDirectory property is checked. If the item is a directory, we need to recurse into that directory. The first step is to create a FileSystemDirectoryReader to handle fetching the directory's contents. That's done by calling the item's createReader() method. Then a new <ul> is created and appended to the parent list; this will contain the directory's contents in the next level down in the list's hierarchy.

After that, directoryReader.readEntries() is called to read in all the entries in the directory. These are each, in turn, passed into a recursive call to scanFiles() to process them. Any of them which are files are inserted into the list; any which are directories are inserted into the list and a new level of the list's hierarchy is added below, and so forth.

Then come the event handlers. First, we prevent the dragover event from being handled by the default handler, so that our drop zone can receive the drop:

dropzone.addEventListener(
  "dragover",
  (event) => {
    event.preventDefault();
  },
  false
);

The event handler that kicks everything off, of course, is the handler for the drop event:

dropzone.addEventListener(
  "drop",
  (event) => {
    let items = event.dataTransfer.items;

    event.preventDefault();
    listing.textContent = "";

    for (let i = 0; i < items.length; i++) {
      let item = items[i].webkitGetAsEntry();

      if (item) {
        scanFiles(item, listing);
      }
    }
  },
  false
);

This fetches the list of DataTransferItem objects representing the items dropped from event.dataTransfer.items. Then we call Event.preventDefault() to prevent the event from being handled further after we're done.

Now it's time to start building the list. First, the list is emptied by setting listing.textContent to be empty. That leaves us with an empty <ul> to begin inserting directory entries into.

Then we iterate over the items in the list of dropped items. For each one, we call its webkitGetAsEntry() method to obtain a FileSystemEntry representing the file. If that's successful, we call scanFiles() to process the item—either by adding it to the list if it's just a file or by adding it and walking down into it if it's a directory.

You can see how this works by trying it out below. Find some files and directories and drag them in, and take a look at the resulting output.

This API has no official W3C or WHATWG specification.

• File and Directory Entries API
• Introduction to the File and Directory Entries API
• DataTransferItem
• FileSystemEntry, FileSystemFileEntry, and FileSystemDirectoryEntry
• Events: dragover and drop